Using Passive Data Collection Methods To Learn Complex Mobility Patterns: An Exploratory Analysis

Collecting detailed information about individual mobility behaviour is a fundamental step in order to implement responsive solutions to mobility issues. One of the biggest challenges is the data collection and extraction of detailed personal travel data, like the type of activity performed on each visited location. As traditional survey methods under represent complex mobility patterns, the current state of the practice is to supplement or eventually replace traditional travel surveys with digital travel surveys and passive data collection methods. Although digital travel surveys can capture complex and heterogeneous user behaviour, they usually require significant inputs from the user, which is supposed to manually complete or confirm part of the survey, hence increasing the respondent's burden when prompted to recall, annotate and classify the activities performed in different locations. The core contribution of this paper is related to the extraction of locations' daily and weekly activity-travel pattern, based on the historical visit patterns. Using raw GPS data, special indexing techniques, and a set of aggregate statistics about activity scheduling and preferences, the proposed methodology provides the probability to classify the activity performed in each location. Results of this exploratory study support the idea that the proposed approach can reconstruct complex mobility patterns while minimizing the number of inputs from the respondent.